Do not worry too much about it at this point. You learned a lot with this piece. It looks excellent, and I would say it would fit in well with an environment. Good work!
Thanks, but I wasn't 100% satisfied until I did it......................I got ndo2 and did this:
Hahah it looks so awesome! (I know it isn't amazing on your level but don't ruin it for me!)haha With ndo2, it was even simpler than I described. I launched ndo2, imported in the cover base texture, clicked one button to convert it to a normal map, messed around with the settings to get the desired effect and boom. I also took it into photoshop and used overlay to increase intensity, it may be a bit too intense, but the important part is, I know how to change it. I didn't intesnify the backcover because I'm lazy like that. :P And yeah, I'm calling this more or less done!
Oh, and without even thinking about it, since I was just testing my models, I automatically made a, what could potentially be an environment. It consists of my crate, the book and a anti-tank gun model I made following Digital Tutors tutorials: Sweet!
I know, obviously in the 3rd model I'm doing, I'll need to consider not having too much of that overlay effect. Talking about my 3rd model, I'm unsure about 2 things. I've labled the two details I'm unsure of in the picture below: 1. Now, how would I create these railings? Splines? You can do it many ways in 3ds max, 2 of which is using the Bend modifier and as stated before, splines. How would you go around doing it? I don't know the best way to do it.
2. Also, how would I create these holes at the top? I don't think an opacity map will work... it has to be geometry right?
Heres my progress so far: Blocking things out atm. Thanks.
If you're looking for an easy way to save your textures in Photoshop, you can take a look at scripts like qSave. There are others mentioned in the wiki.
Or you can still create actions for that yourself For the easiest one: New action -> Flatten image -> Save As .tga -> Undo (get your layers back) -> Save (if you want to save your PSD at the same time) -> Stop recording, then maybe assign a keyboard shortcut to it. You can of course get fancier if you want to save specific textures in specific channels, but then those scripts might be easier to use
I see you already chose nDo2 for your normal map conversion / creation. If you ever decide to go the Substance Designer / Painter route, they also have Bitmap2Material which can help you achieve the same kind of things (at least the conversion, the creation from scratch can be done in SD/SP).
You can of course use both, whichever your are the most comfortable with. Have fun!
The railing I would create with a bend modifier, splines would work as well it truly depends on what you think will work best.
The holes probably should be geo, though I am not even sure what the hell they are. They read a bit oddly. Could be holes, could be small supports, could be pegs.
Thanks a ton! I will definitely check those scripts out! And I'm planning to use both NDO2 and substance designer/painter. NDO for normals and SD/SP for its texture creation glory.
Hey guys, I'm having topology problems. So I'm making the highpoly version, and this needs to be all quads since sub-div smoothing doesn't handle triangles well obviously.. but I can't figure out a way to quadrify the mesh in a clean way. TURBOSMOOTH:
I think the issue here is you don't fully understand how SubDiv modeling works. And thats okay. here is something that will help you out abit. This to.
Thanks a lot. Most of that I knew but it definitely shed some light sub div modelling. Unfortunately that doesn't directly help me out, I think I'll have another go at trying to create better topology for the model. Any and all help would be very well appreciated.
Ohh! I totally forgot about Millennias tutorials! I'll make a start on that today then. I will also work on this at the side. Also, thanks for the helpful instructions.
Thanks to your guidance, I got some good looking results: Now, I did get 2 artifacts on the base of the model, that however is not from smoothing, it was there even before smoothing, it's my vertices, they're not completely aligned. I could, but I'm going to leave it since it's a minor issue. I'll strive to fix that kind of stuff in my next models as I progress further. Now the scary part. Make the lowpoly by deleting all un-necessary edges... oh the amount of problems I'm going to get in my normal and ao map ;/
Okay so how would I do this? I have objects that are clipping, for exmaple, the 2 wooden blocks nailed on the window... so do I explode everything? Or do I just make the lowpoly, non-exploded and just make a cage out of that? So far the base of the model wasn't very hard to make a lowpoly out of. Simply deleted turbosmooth, deleted all support edges. Everything else should be easier so.... green light so far.
Yeah same deal, explode once done. You can keep a version exported with the meshes closer together for better Ambient occlusion but for Normal map baking, 100% explode the mesh.
Your railing looks like it needs to be a bit beefed up, looks a tad thin. Those artifacts you have are most likely caused by severely tight edges or the face is not 100% flat/planar.
How are you going about adding the wood trim around the door, window, and around the shed?
Understood. Yeah, thanks for noticing, I will beef up my railing...oh god thats sounds weird. Hahah
About the trim... I'll be honest. I don't know. I thought I could use the faces and paint them white, but I'll have to move my geometry around now and scale some faces up... do you know any other way?
Also, for normal maps, does the mesh 100% need to be exploded? Every single time? Just making sure.
One more thing, I forgot about the holes at the top. I'm leaving it, I'll make a note of remembering every detail next time. One step at a time.
Note: I also want to incorporate Substance Painter AND marmoset into my workflow. I want good presentation and quality texturing so it's kind of essential I get the software for it.
Yes, if you use xNormal you need to explode your mesh every single time for normal map baking, not for AO baking, this depends. Sometimes I bake three times for AO. One for all meshes to be on top of each other to get real Ambient Occlusion, then bake exploded to get AO on itself, then once again for CAO (Crevice Ambient Occlusion). CAO is another bake option which allows you the artist a better control on tighter areas that normal AO would not accurately read (crevices). Usually my workflow consists of baking the Normal and AO in xNormal and the swath of remaining bakes in Substance Painter. I just like xNormal's result a tad better, painter does a damn close job as well, faster at that.
Marmoset is a necessity, as is Substance though, if you are strapped for cash Substance has a very nice renderer now as well. I prefer Marmoset because of its ease of use and utility, Marmoset 3 should be announced soon, perhaps it is wise to hold off on that program until then.
While substance painter is good, i would recommend you learning the old fashion Photoshop way to texture. Maybe mix both.
In some ways I agree with you, however, I would think it is better for @CodeferBlue to get a strong grasp on current methods, now. After that process perhaps a return to traditional methods may be in order. It is a tougher process compared to current pipelines.
Ah, okay, understood. And yeah I like the idea of learning current gen techniques now, and then older later. I actually want to learn both, it will only increase my capabilities more. It may be wise to hold off on toolbag2 since toolbag3 may come out soon... But.. I'm still eager to try it out so I'll get it.
also, what should I do about the trims? Any way I could create those without messing with my topology?
Ah yeah forgot to comment on that, I would do that in the geo.
It will mess your topology... honestly you could try and add the trim as floating geo. Just make the pieces and place them around the doorway and window.
It would be better to cut it into the actual shed rather than floating, up to you.
Understood, I could have done it in the topology, but I was too depressed over the fact that I couldn't figure out how to make quads. Again, making notes of things like this so it doesn't happen to my next models.
and yeah it will cut into the mesh. It will be a very slight, thin piece. I'll work on it tomorrow since it's kind of late where I'm at. I'm at the other side, it must be around 6ish or 7ish where you are haha.
hey guys, I'm 3/4 of the way through unwrapping my model. I'm wondering about UV channels. I want different aspects of my model to be on different channels. MY maps are filling up quick, I'm making sure they're packed in tight and nicely but they also have good resolution. I'm also thinking out texture priority. Should I have 2 UV channels? I think I will most defo need 2. 1 for the base part of the model, the structure. The 2nd will be for everything else (roof, stairs). Any drawbacks from using 2 UV channels?
You should not need 2 UV channels for this piece. Just use a 4096 texture it will be more than enough.
Two UV sets for one piece like this would be Albedo/Diffuse UV set and the second would be your lightmap UV set.
If you really want a second material set for something, usually that is done for a hero prop/mesh that has multiple materials metal, leather, plastic, etc.
Really? hm. See, this is the thing, I don't know properly know when to use multiple channels, or, how much resolution should I use? These factors, I have no clue of. I mean, I know the more resolution I use the higher the quality my textures will be, but does resolution have anything to do with UV map size?
Also, use a checker pattern if you aren't, it helps alot with UV's.
I do use a checker pattern. But, I'm unsure of something. See, what I do is this: I scale up the UV shells I'm working on to get a lot of checker density on them, so with a fair amount of squares, I clearly see and fix distortion and keep it at a minimal, I then scale them down to fit them into the 0,1 UV space. At this point, I obviously don't get many squares, this is where I'm confused at. When placing them into 0,1 uv space, does the amount of squares matter? Or is it the actual texture resolution, that matters when I render out my UV template for texturing?
My plan is, all the trims can use one texture sheet, since they all look the same. So I could just get a white noisy, painted texture and I can use that as the trim. But I don't know if I'll get enough resolution out of that. I don't really have a proper idea of how big somethings in my UV map should be. Mentally, I'm preparing my self to not get so depressed over not knowing what to do, I'm going to embrace the challenge and learn to overcome it and learn from it, so that my next model, I don't run into problems like this. Progress after progress, I'm going to filter out these problems, fill in the gaps of knowledge that I don't have yet, this will help me make models like this with a breeze.
So, this can get a bit more complex than you probably need at the moment. I will explain some basics and a bit more... advanced I suppose.
Most if not all objects in a game will have two UV sets. What are these UV sets?
UV set one = This is your diffuse map/albedo map UVs. Here all the UVs that will display your textures are stored, Albedo, normal, metalness, roughness, AO, emissive, height, etc. These UVs can exit the 0 to 1 space for tiling textures (more on that later).
UV set two = This stores alternate UVs that store the UVs for LightMap data. These UVs are used in game engines to properly bake light/shadows onto your model. This can be a completely different set of UVs that must not be flipped, contain clipping, or contain errors. All edges that can be sewn should be sewn to remove seams. These UVs cannot exit the 0 to 1 space.
UV sets past this = Usually used for tiling emissive maps, or animating some type of texture without relying on properly unwrapped UVs like is set 1. Also used for layering an alternate texture over the base texture. Some outlier methods/techniques and some texture artist magic I am not intimate with.
Static meshes that do not move will need a lightmap, skeletal meshes which do move for the large most part do not need a lightmap, as they are moving and require real lighting.
When you are texturing an object and are not tiling their textures you must stay in the the UV space at all times! Stay in the 0 to 1 space. If you are tiling a material/texture you can have that UV hull/piece/object wrap out side of the 0 to 1 space to get optimal resolution without doing so in the texture but rather in the UVs. Pipes, wires, cords, modular walls/floors, terrain, detail maps are examples of this, really any place that a tiling texture would make sense.
Determining resolution based upon your UVs is determined by what area of the piece is most important? If I am making a character the face is the most important, this should get a better portion of the UV space per a ratio basis, that is to say if the whole character is using the same material and/or UV set. What most artists do if they are working with a hero prop/character will have multiple materials for that object. So, for a character they could have literally 2, 3, 5, or much more materials applied with unique UVs for that hero object. A gun for example would/could have the gun, magazine, scope, and other attachments on different UV sets. Call of Duty or Battelfield is doing this, if not drastically more separated.
Specifically regarding your UVs and your shed piece, You can if you really want to have two UV sets, but honestly that is not needed. I would fit everything on the one UV set in one 0 to 1 space. You really need to jam pack that UV space to capacity, rotate, flip and maneuver each UV hull to make use of the space available. Also, straighten the board UVs to make them more, efficient. The shed, trim, roof, and steps can all fit.
The scale of the checkerboard pattern is irrelevant to the resolution of your UVs. You can pump the scale and tile quantity of the checkerboard pattern in the material within Max.
UVs are there to ensure the data is accurately translated for the game engine to project your textures. If the UVs are shit and you have great textures the result will be shit. If you have Excellent UVs but shit textures, the result will be shit. If all is excellent but you have small textures it will be blurry, pixelated, and look bad. Optimal UVs require optimal textures based upon your needs or the specifications of the project. You could, could fit a whole entire building in one UV set but would need a larger texture resolution. Just some ways to look at it.
Okay. I see. Few things I'm concerned about. 1. "A gun for example would/could have the gun, magazine, scope, and other attachments on different UV sets.". So I could have props, where different parts of the prop could be on different UV channels? Any drawbacks?
2. "Also, straighten the board UVs to make them more, efficient" never understood what you meant there.
Also, what is a lightmap? Secondly, will I need to create a second UV channel for it to be needed if I want optimal rendering in UE4? In other words, do I need it?
This is becoming a whole lot clear now. After reading your comment, I've realised that my uv shells are way out of scale. They should be smaller, on a 4096 resolution map. I'm still trying to calibrate myself through this stage of 'will I have enough resolution on my textures or not?'. The more of this kind of work I do, it's get an instinct for resolution and texture density.
ahh, I've ran into a problem. I converted my highpoly mesh into an editable poly. So the turbosmooth modifier is applied onto the physical properties of the mesh. I'm going to have to create all of it except for the shed again. Not hard but tedious. To make the lowpoly mesh, I delete turbosmooth and the supporting edges = LP Mesh.
1.) The largest drawback is higher cost to the game. Higher cost, mainly. More textures, more materials, more memory taken.
2.) make your trim/wood board uvs straighter, not crooked. If you make them straighter they will fit in the uv space better.
A light map is a texture that is created in a game engine using you second uv set as a projection. Light maps bake lighting onto your model, shadows mainly. if you want this in unreal it would be good practice to make a light map. The light map is in the second uv channel, it's not super necessary for one prop but for an environment it would be.
Hey guys, don't worry, I'm not dead I've just been taking a long time doing the UV unwrappinf. I'm 95% done, I'm just having trouble UV unwrapping the roof. Things I've tried: Planar Projection Z Axis. This works okay but since not all the face are at a flat angle, distortion is pretty frequent. Flatten Mapping. This basically takes every face and breaks them apart, leaving me to relax each individual face and then stich them back together. Ultimately, this would give me what I want, but the process of doing it would taking a long time and would be highly tedious. How would you guys UV unwrap this kind of geometry?
BTW, I can't wait to get to the texturing stage. I know I'm gonna have fun in painter, as long as I can figure out how to use it.
Hey guys, don't worry, I'm not dead. I've just been taking a long time UV unwrapping. I'm 95% done, I just can't figure out how to UV unwrap the roof. Things I've tried so far: Planar Projection Z Axis: This works okay, but since every face isn't at a straight angle, distortion is quite frequent. Flatten Mapping: This essentially takes every face of the mesh, and breaks them apart. Leaving me with having to relax each and individual face and stich them all back up together. Ultimately, this would give me the desired end result, but the process of doing this is very tedious and not to mention it would take a while. So how would I UV unwrap this? How would you guys do it?
BTW, I can't wait to get onto the texturing stage. It's gonna be pretty fun in Substance Painter provided I can figure out how to use it.
I would do a planar projection from Y. Then with the resulting UVs selected I would hit unfold, Vertical then horizontal. Then because they are all hard edges, I would separate each hull.
Isnt your roof overtly sharp? Dont you want it more sloped and rounded metal?
Hm, okay. Cept' I don't know what you meant by this "unfold, Vertical then horizontal. ". What would the 3ds max equivelant be for that? Don't completely understand.
Hmm, I guess my roof is kind of sharp, but what would I slope? I would probably smooth and slope out the top edges of the roof, but the border edges, I wouldn't really want to slope/smooth those out right?
Also, I would like your recommendation on something. Wacom tablets. I'm going to need one in the near future, for many things. Concept, texturing (I will be creating a lot of stylized textures amongst many other 2d related things). I'm sure you know a Wacom Tablet is treasured equipment in a 3d artists toolbox. So which one would you recommend? I'll most likely create a thread for this so I can get many opinions and recommendations. Like I stated before, I need it for concept art, texturing and no doubt other things too. Currently, I just want to see which one would be good for me and how much they are going for.
Okay. The iron tool isn't a 3ds max deafult tool. It's part of a set of UV mapping tools all tucked neatly within a plugin called Tex Tools. It's really good, so I'll get it and see what I can do.
Also, yeah you're right about the thickness of the roof, but don't ya think the peaks and valleys are a bit too smooth in that picture? In the concept, the roof is very... raw, sharp, non refined. I will defo add some thickness, but I'm not really gonna smooth it out.
And thanks for your recommendation, interesting. The bamboo is a defo no. I've heard from a few people that it's less capable and its more of a childs toy. Hence why it's cheaper, making people fall for it's price is a big mistake. The Intuos models are what the pros use, and they are an industry standard. Now to find out, which Intuos? I'll start a thread for it. Thanks though.
Ohh! I nearly forgot! So I was kind of curious, so I typed in on Google: 'game art 3ds max'. I found this:
Isn't that cool? I know the very first area of the course is pretty basic so I already know that stuff, but quarter way in, it gets pretty good in both courses! Especially, it teaches me workflows AND substance painter as well as marmoset. Not only that, but thankfully the courses are incredibly recent. So accuracy is not a doubt. Whaddya think?
Yeah, what little I have used of 3ds max I immediately downloaded textools.
Its up to you how you want to make the roof, a concept is not always followed 100%, even in a professional setting artists can deviate in some cases.
The bamboo is excellent if you are strapped for cash but compared to the intuos 5 its a farcry. Nothing is better than a cintiq but... thats a pretty penny.
Those courses look interesting, I cannot speak to their worth. I am sure you would learn from them regardless.
About the cintiq vs intuos, are there any actual functional changes between them? I know the Cintiq has you drawing on the actual tablet, but is there anything a Cintiq can do an Intuos can't? I don't really want a Cintiq. Not because it's price is too high, but I wouldn't need it. An intuos tablet that I can use to accurately create nice concepts and texture work from is perfectly fine. I'm not a professional 2d drawing type of artist. I just need something capable at the side to make handpainted artwork from that can aid my 3d work.
Yeah those courses look amazing. I'll take them. I don't know about the second one, maybe in the future, but the first one still has a lot to teach.
Well the cintiq allows you to actually paint/brush/sculpt directly on an image/screen which is excellent. There are little functionality differences between the two (cintiq and intuos) however I feel it is more accurate just not significantly.
Replies
Do not worry too much about it at this point. You learned a lot with this piece. It looks excellent, and I would say it would fit in well with an environment. Good work!
Move on, and learn more.
Hahah it looks so awesome! (I know it isn't amazing on your level but don't ruin it for me!)haha
With ndo2, it was even simpler than I described. I launched ndo2, imported in the cover base texture, clicked one button to convert it to a normal map, messed around with the settings to get the desired effect and boom.
I also took it into photoshop and used overlay to increase intensity, it may be a bit too intense, but the important part is, I know how to change it. I didn't intesnify the backcover because I'm lazy like that. :P
And yeah, I'm calling this more or less done!
Sweet!
NDO is awesome, but make sure you have that normal map detail set to Overlay on your bake base normal map.
Nice gun man!
Talking about my 3rd model, I'm unsure about 2 things. I've labled the two details I'm unsure of in the picture below:
1. Now, how would I create these railings? Splines? You can do it many ways in 3ds max, 2 of which is using the Bend modifier and as stated before, splines. How would you go around doing it? I don't know the best way to do it.
2. Also, how would I create these holes at the top? I don't think an opacity map will work... it has to be geometry right?
Heres my progress so far:
Blocking things out atm.
Thanks.
If you're looking for an easy way to save your textures in Photoshop, you can take a look at scripts like qSave.
There are others mentioned in the wiki.
Or you can still create actions for that yourself
For the easiest one: New action -> Flatten image -> Save As .tga -> Undo (get your layers back) -> Save (if you want to save your PSD at the same time) -> Stop recording, then maybe assign a keyboard shortcut to it.
You can of course get fancier if you want to save specific textures in specific channels, but then those scripts might be easier to use
I see you already chose nDo2 for your normal map conversion / creation. If you ever decide to go the Substance Designer / Painter route, they also have Bitmap2Material which can help you achieve the same kind of things (at least the conversion, the creation from scratch can be done in SD/SP).
You can of course use both, whichever your are the most comfortable with.
Have fun!
Just will be tedious to do it all the way around.
The holes probably should be geo, though I am not even sure what the hell they are. They read a bit oddly. Could be holes, could be small supports, could be pegs.
thanks for the help!
TURBOSMOOTH:
I think the issue here is you don't fully understand how SubDiv modeling works. And thats okay.
here is something that will help you out abit. This to.
He goes over sub div modeling and talks about support edges.
To retain quads, delete all edges that are not defining the door, window, etc. Add edge loops near each vert. Merge the edge with the vert nearest.
Add edges to support definition and retain overall shape and smoothing.
Red Xs = Delete edge
Deep Blue = Add new edges for quads
Light Blue = Support edges
Might have missed some edges but that should give you a good idea.
I'll make a start on that today then. I will also work on this at the side. Also, thanks for the helpful instructions.
Now, I did get 2 artifacts on the base of the model, that however is not from smoothing, it was there even before smoothing, it's my vertices, they're not completely aligned. I could, but I'm going to leave it since it's a minor issue. I'll strive to fix that kind of stuff in my next models as I progress further.
Now the scary part. Make the lowpoly by deleting all un-necessary edges... oh the amount of problems I'm going to get in my normal and ao map ;/
Your railing looks like it needs to be a bit beefed up, looks a tad thin. Those artifacts you have are most likely caused by severely tight edges or the face is not 100% flat/planar.
How are you going about adding the wood trim around the door, window, and around the shed?
About the trim... I'll be honest. I don't know. I thought I could use the faces and paint them white, but I'll have to move my geometry around now and scale some faces up... do you know any other way?
Also, for normal maps, does the mesh 100% need to be exploded? Every single time? Just making sure.
One more thing, I forgot about the holes at the top.
Note: I also want to incorporate Substance Painter AND marmoset into my workflow. I want good presentation and quality texturing so it's kind of essential I get the software for it.
Thanks.
Maybe mix both.
Marmoset is a necessity, as is Substance though, if you are strapped for cash Substance has a very nice renderer now as well. I prefer Marmoset because of its ease of use and utility, Marmoset 3 should be announced soon, perhaps it is wise to hold off on that program until then.
In some ways I agree with you, however, I would think it is better for @CodeferBlue to get a strong grasp on current methods, now. After that process perhaps a return to traditional methods may be in order. It is a tougher process compared to current pipelines.
also, what should I do about the trims? Any way I could create those without messing with my topology?
It will mess your topology... honestly you could try and add the trim as floating geo. Just make the pieces and place them around the doorway and window.
It would be better to cut it into the actual shed rather than floating, up to you.
and yeah it will cut into the mesh. It will be a very slight, thin piece. I'll work on it tomorrow since it's kind of late where I'm at. I'm at the other side, it must be around 6ish or 7ish where you are haha.
Two UV sets for one piece like this would be Albedo/Diffuse UV set and the second would be your lightmap UV set.
If you really want a second material set for something, usually that is done for a hero prop/mesh that has multiple materials metal, leather, plastic, etc.
You should be totally 100% fine with one UV set.
I do use a checker pattern. But, I'm unsure of something. See, what I do is this: I scale up the UV shells I'm working on to get a lot of checker density on them, so with a fair amount of squares, I clearly see and fix distortion and keep it at a minimal, I then scale them down to fit them into the 0,1 UV space. At this point, I obviously don't get many squares, this is where I'm confused at. When placing them into 0,1 uv space, does the amount of squares matter? Or is it the actual texture resolution, that matters when I render out my UV template for texturing?
My plan is, all the trims can use one texture sheet, since they all look the same. So I could just get a white noisy, painted texture and I can use that as the trim. But I don't know if I'll get enough resolution out of that. I don't really have a proper idea of how big somethings in my UV map should be.
Mentally, I'm preparing my self to not get so depressed over not knowing what to do, I'm going to embrace the challenge and learn to overcome it and learn from it, so that my next model, I don't run into problems like this. Progress after progress, I'm going to filter out these problems, fill in the gaps of knowledge that I don't have yet, this will help me make models like this with a breeze.
Most if not all objects in a game will have two UV sets. What are these UV sets?
- UV sets past this = Usually used for tiling emissive maps, or animating some type of texture without relying on properly unwrapped UVs like is set 1. Also used for layering an alternate texture over the base texture. Some outlier methods/techniques and some texture artist magic I am not intimate with.
Static meshes that do not move will need a lightmap, skeletal meshes which do move for the large most part do not need a lightmap, as they are moving and require real lighting.When you are texturing an object and are not tiling their textures you must stay in the the UV space at all times! Stay in the 0 to 1 space. If you are tiling a material/texture you can have that UV hull/piece/object wrap out side of the 0 to 1 space to get optimal resolution without doing so in the texture but rather in the UVs. Pipes, wires, cords, modular walls/floors, terrain, detail maps are examples of this, really any place that a tiling texture would make sense.
Determining resolution based upon your UVs is determined by what area of the piece is most important? If I am making a character the face is the most important, this should get a better portion of the UV space per a ratio basis, that is to say if the whole character is using the same material and/or UV set. What most artists do if they are working with a hero prop/character will have multiple materials for that object. So, for a character they could have literally 2, 3, 5, or much more materials applied with unique UVs for that hero object. A gun for example would/could have the gun, magazine, scope, and other attachments on different UV sets. Call of Duty or Battelfield is doing this, if not drastically more separated.
Specifically regarding your UVs and your shed piece, You can if you really want to have two UV sets, but honestly that is not needed. I would fit everything on the one UV set in one 0 to 1 space. You really need to jam pack that UV space to capacity, rotate, flip and maneuver each UV hull to make use of the space available. Also, straighten the board UVs to make them more, efficient. The shed, trim, roof, and steps can all fit.
The scale of the checkerboard pattern is irrelevant to the resolution of your UVs. You can pump the scale and tile quantity of the checkerboard pattern in the material within Max.
UVs are there to ensure the data is accurately translated for the game engine to project your textures. If the UVs are shit and you have great textures the result will be shit. If you have Excellent UVs but shit textures, the result will be shit. If all is excellent but you have small textures it will be blurry, pixelated, and look bad. Optimal UVs require optimal textures based upon your needs or the specifications of the project. You could, could fit a whole entire building in one UV set but would need a larger texture resolution. Just some ways to look at it.
1. "A gun for example would/could have the gun, magazine, scope, and other attachments on different UV sets.". So I could have props, where different parts of the prop could be on different UV channels? Any drawbacks?
2. "Also, straighten the board UVs to make them more, efficient" never understood what you meant there.
Also, what is a lightmap? Secondly, will I need to create a second UV channel for it to be needed if I want optimal rendering in UE4? In other words, do I need it?
This is becoming a whole lot clear now. After reading your comment, I've realised that my uv shells are way out of scale. They should be smaller, on a 4096 resolution map. I'm still trying to calibrate myself through this stage of 'will I have enough resolution on my textures or not?'. The more of this kind of work I do, it's get an instinct for resolution and texture density.
Appreciate the help as always.
To make the lowpoly mesh, I delete turbosmooth and the supporting edges = LP Mesh.
2.) make your trim/wood board uvs straighter, not crooked. If you make them straighter they will fit in the uv space better.
A light map is a texture that is created in a game engine using you second uv set as a projection. Light maps bake lighting onto your model, shadows mainly. if you want this in unreal it would be good practice to make a light map. The light map is in the second uv channel, it's not super necessary for one prop but for an environment it would be.
Flatten Mapping. This basically takes every face and breaks them apart, leaving me to relax each individual face and then stich them back together. Ultimately, this would give me what I want, but the process of doing it would taking a long time and would be highly tedious. How would you guys UV unwrap this kind of geometry?
BTW, I can't wait to get to the texturing stage. I know I'm gonna have fun in painter, as long as I can figure out how to use it.
I've just been taking a long time UV unwrapping. I'm 95% done, I just can't figure out how to UV unwrap the roof. Things I've tried so far:
Planar Projection Z Axis: This works okay, but since every face isn't at a straight angle, distortion is quite frequent.
Flatten Mapping: This essentially takes every face of the mesh, and breaks them apart. Leaving me with having to relax each and individual face and stich them all back up together. Ultimately, this would give me the desired end result, but the process of doing this is very tedious and not to mention it would take a while. So how would I UV unwrap this? How would you guys do it?
BTW, I can't wait to get onto the texturing stage. It's gonna be pretty fun in Substance Painter provided I can figure out how to use it.
Isnt your roof overtly sharp? Dont you want it more sloped and rounded metal?
Don't completely understand.
Hmm, I guess my roof is kind of sharp, but what would I slope? I would probably smooth and slope out the top edges of the roof, but the border edges, I wouldn't really want to slope/smooth those out right?
Also, I would like your recommendation on something. Wacom tablets. I'm going to need one in the near future, for many things. Concept, texturing (I will be creating a lot of stylized textures amongst many other 2d related things). I'm sure you know a Wacom Tablet is treasured equipment in a 3d artists toolbox. So which one would you recommend? I'll most likely create a thread for this so I can get many opinions and recommendations. Like I stated before, I need it for concept art, texturing and no doubt other things too. Currently, I just want to see which one would be good for me and how much they are going for.
Thicken the roof, round out the peaks and valleys:
Regarding tablets, ehh go with an intuos 5 small if you have the money, thats what I started with.
http://www.amazon.com/Wacom-Intuos5-Medium-Tablet-PTH650/dp/B0076HMDSC
or a bamboo wacom, bit cheaper
http://www.amazon.com/Wacom-Bamboo-CTL471-Tablet-Black/dp/B00EVOXM3S/ref=sr_1_6?s=pc&ie=UTF8&qid=1460734238&sr=1-6&keywords=Wacom
Also, yeah you're right about the thickness of the roof, but don't ya think the peaks and valleys are a bit too smooth in that picture? In the concept, the roof is very... raw, sharp, non refined. I will defo add some thickness, but I'm not really gonna smooth it out.
And thanks for your recommendation, interesting. The bamboo is a defo no. I've heard from a few people that it's less capable and its more of a childs toy. Hence why it's cheaper, making people fall for it's price is a big mistake. The Intuos models are what the pros use, and they are an industry standard. Now to find out, which Intuos? I'll start a thread for it. Thanks though.
Ohh! I nearly forgot! So I was kind of curious, so I typed in on Google: 'game art 3ds max'. I found this:
http://gameartunboxed.com/course/intro-3d-modeling-games-3ds-max/
And this:
http://gameartunboxed.com/course/masterclass-modeling-props-for-aaa-games/
Isn't that cool? I know the very first area of the course is pretty basic so I already know that stuff, but quarter way in, it gets pretty good in both courses! Especially, it teaches me workflows AND substance painter as well as marmoset. Not only that, but thankfully the courses are incredibly recent. So accuracy is not a doubt. Whaddya think?
Its up to you how you want to make the roof, a concept is not always followed 100%, even in a professional setting artists can deviate in some cases.
The bamboo is excellent if you are strapped for cash but compared to the intuos 5 its a farcry. Nothing is better than a cintiq but... thats a pretty penny.
Those courses look interesting, I cannot speak to their worth. I am sure you would learn from them regardless.
I see, I'll think about the roof.
About the cintiq vs intuos, are there any actual functional changes between them? I know the Cintiq has you drawing on the actual tablet, but is there anything a Cintiq can do an Intuos can't? I don't really want a Cintiq. Not because it's price is too high, but I wouldn't need it. An intuos tablet that I can use to accurately create nice concepts and texture work from is perfectly fine. I'm not a professional 2d drawing type of artist. I just need something capable at the side to make handpainted artwork from that can aid my 3d work.
Yeah those courses look amazing. I'll take them. I don't know about the second one, maybe in the future, but the first one still has a lot to teach.